CLIMB: 基于质心的层次记忆用于在线持续自监督学习
CLIMB: Centroid-Based Hierarchical Memory for Online Continual Self-Supervised Learning
AI总结：
提出CLIMB方法，结合层次质心记忆和知识蒸馏，在内存受限的在线无标签数据流中学习表示，优于现有OCSSL方法。
摘要：
在线持续自监督学习（OCSSL）旨在从连续的无标签数据流中学习表示，无需任务边界知识且受内存限制。现有方法要么依赖利用潜在空间结构的重放缓冲区，要么仅依赖正则化。我们提出CLIMB（持续学习与智能记忆库），它同时结合了这两者。我们的方法引入了一种基于质心的层次记忆，存储图像总数有界，并结合对重放示例的知识蒸馏以限制表示漂移。该记忆将相似图像分组为质心，为对比学习提供难以区分的示例，同时覆盖观测分布的多样性。在Split CIFAR-100和Split ImageNet-100上的实验，包括现有最先进的标准基准以及一个具有不规则任务分布的新协议，表明CLIMB优于最先进的OCSSL方法。
英文摘要：
Online Continual Self-Supervised Learning (OCSSL) aims to learn representations from a continuous stream of unlabeled data, without knowledge of task boundaries and under memory constraints. Existing methods rely either on replay buffers that exploit latent space structure, or on regularization alone. We present CLIMB (Continual Learning with Intelligent Memory Bank), which combines both simultaneously. Our method introduces a hierarchical centroid-based memory, bounded in total number of stored images, combined with knowledge distillation on replayed examples to limit representation drift. The memory groups similar images into centroids, providing hard-to-discriminate examples for contrastive learning while covering the diversity of observed distributions. Experiments on Split CIFAR-100 and Split ImageNet-100, on standard benchmarks from the state-of-the-art as well as a new protocol with irregular task distributions show that CLIMB outperforms state-of-the-art OCSSL methods.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
